Interactive comment on “ Supraglacial debris cover assessment in the Caucasus Mountains , 1986-2000-2014

General comments: A thorough language editing is needed as there are issues and awkwardness in the sentence structure right from the first sentence of the abstract. I will not be pointing out them separately as they are present in almost every other sentence. There are several issues with the methods too as highlighted by Alifu, H. in his interactive comments. In the light of these issues, presently it is really difficult to comment upon the accuracy of the results. Also, how is this a review article as I can see in the manuscript type? It certainly is a research article and not a comprehensive review.


Introduction
The introduction lacks focus, despite bringing in some background information about the importance of glaciers.For ex. the issue of debris thickness is mentioned but the objectives of the paper do not target debris cover thickness.The introduction would need to be revised/focused.Most of the statements need to be expanded since they do not fully explain the gaps in knowedge, and remain vague, for example p 2 lines 31 to 37: "The Greater Caucasus SDC is an important control for ice ablation, as it is similarly in many other glaciated areas (Lambrecht et al., 2011) and has been identified as a key player in glacier mass balance (Popovnin and Rozova, 2002)." This would need to be expanded and the concepts explained.
Similarly: "In addition, in some cases SDC and proglacial lakes are directly related to glacial hazards" This remains vague and issues are mentioned but not developed.

Study area and previous studies
The different parts of the range are mentioned but it is not clear how these were defined.The 1st paragraph on p.4 is indented to be a literature review but it mixes glacier area change with debris cover mapping, these are not clearly addressed.The authors could list the studies that determined area changes, the extent of the areas and the remaining gaps; then the debris cover mapping issues could be addressed separately.P4 l.12-15 "This research further aims to compare glacier mapping using manual and semi-automated methods to assess the SDC change across a larger area than previous studies according to the western, central and eastern Greater Caucasus over the last 30 years." This paragraph sounds like it belongs to introduction, not study area/previous studies.
Also, here the objectives do not come out clearly; while in the abstract and introduction the author mention the need to document glacier area changes, here this points to another objective, which is to compare various methods to delineate debris cover.
Overall I suggest that the authors sharpen their objective and focus on glacier area changes in the various parts of the ranges, perhaps splitting the glaciers into clean and debris cover glaciers.To achieve this objective, the authors could just choose their best method for delineating debris cover and provide an uncertainty analysis.As C3 written currently, this almost contains two different studies and it prevents this from developing either.

Data and methods
Sections 3.2 to 3.4 summarize various methods for delineating debris cover, and it seems to be a literature review.This is fine for a paper that would focus on comparing various methods, but here the objective seems to be different and this much detail might not be needed.Sections 3.2 to 3.4 could be combined and shortened, and the authors could just present the approach they took in a section on glacier clean ice and debris cover delineation, for example.Furthermore: the method tested is presented for the 2013 image, and it's not clear if the same method was applied also to the other images.These sections can be clarified.P5 l 13-16 : I do not see the use of presenting the Alifu et al 2015 methodology here, if it is shown later (in results) that their method did not work for this study.I suggest removing the comparison with the Alifu paper in this section as well as in results (p.9 l .16-19).Again, this distracts the reader from the main goal of the paper which gets buried into the details of SDC mapping.Section 3.5 on comparison with manual digitization can also be presented as part of the glacier mapping section.The SPOT imagery used to ]check the SDC tongues is from 2016, but the Landsat imagery was from 2014-are the authors certain that there was no change in the glacier extents?Area changes might be small but I am not sure this can be used as a check for the classification if the images are not from the same year..

Results and discussion
section 4.1 Supraglacial debris cover (SDC) assessment using Thermal/Near-IR/Mid-IR band ratio methodology Again the focus here is not on assessing the accuracy of the delineation methodology but on the changes in debris cover, it seems-in this case it is unnecessary to mention the method used in the section title.
p. 9 l 16-25: see my comments above about Alifu et al paper.I suggest removing this comparison and just focusing on the method presented, it the authors are confident in it.
Fig. 4 in the manual outlines, it is not clear if the NW part of the glacier tongue delineated as debris cover is supraglacier debris or just moraine.Have the authors checked this area with the high-resolution imagery?I suggest that in fig 4 it would be more helpful to present the high-resolution imagery rather than the Alifu et al outlines.P10 l 4-5 "Therefore, the Alifu et al. methodology to identify debris covered ice in the Greater Caucasus cannot be 5 considered robust for OLI imagery, and unsuitable for extended time series".Again the evaluation of Alifu et al method is beyond the scope of the paper as I understand it, and the authors would benefit from focusing on their own method here.P10 section 4.2 and 4.3 similar to my comments about the methodology, the paper wonders about here with the comparison of the different methods.I think one method should be chosen over the other and the results of that classification method should be chosen.
Table 2: -total glacier number should be "total number of glaciers" -glacier number should be "number of glaciers" since the authors do not refer to glacier IDs -"debris cover should be "debris covered ice" Figure 6 does not tell much in its present form, sicne there is almost no change.This could be just mentioned in the text, with the C5 Results on SDC p. 11-12 are hard to follow, and in some parts are descriptive, some ways of improving include: -Separating the glaciers only by range rather that range and glacier area.For ex, in figures 7 and 8 it is hard to distinguish any difference among the rages since different colors are used for the same class (DC) -Clarifying what "percent" the authors are referring to-sometimes this relates to -It would help if instead of average slope, the authors did a correlation between slope of each glacier and its -The large changes (+49 Fig 11 a: The multi-temporal glacier mapping extents shown here are questionable, looking at the area covered by the nutatak.It seems that a) either a lot of the nuatak was exposed in 2016 by surface lowering, or b) that the 1986 image has a lot of snow which covered some of the area around the nutatak.In this case, the area changes are affected by the way that the nunataks were mapped.
For consistency, a number of authors chose to consider the area of the nunataks constant and only estimate the area changes due to glacier retreat.This issue should be carefully assessed since it can be introduce large errors in the change estimates.the glacier outlines overlaid on the SPOT imagery do not seem to much; it is hard to see but it seems like the SDC area is over-estimated (unless there was a large area change sine 2014 which is unlikely).Also I am not sure how the GPR analysis helps here, it seems to have been added to the paper since the measurements were taken; however the purpose of the paper is not to provide ice thickness measurements; The GPR section seems to be out of the purpose of the paper and can be removed/saved for a future paper on glacier thickness.
Fig 13: looking at the areas of increase in DC, it seems that some of the "apparent" increase is indeed due to difference on mapping.For ex fig 13b shows an increase in DC area in the glacier accumulation area, which seems to be rather due to the way that the rock outcrops were mapped.Similarly there is a large increased where two glacier tongues marge-this can be a similar case of mapping differences.

Fig. 3
Fig.3 seems to belong to the results section, or can be presented as part of the uncertainty analysis.

Fig
Fig 12:the glacier outlines overlaid on the SPOT imagery do not seem to much; it is hard to see but it seems like the SDC area is over-estimated (unless there was a large area change sine 2014 which is unlikely).